Phoebe Ellerwick

Chief Product Officer / Australia

PhoebeEllerwick

Portrait of Phoebe Ellerwick

A household product should fit the life around it, not require that life to fit the catalogue.

I joined the fictional Cervellan Homewares in February 1994, shortly after completing industrial design at Ternovale Design College.

Open to
  • Board appointments
  • Speaking invitations
The full storyTalk to me

A household product should fit the life around it, not require that life to fit the catalogue.

I develop household ranges by looking beyond the object itself. Cupboard depth, assembly instructions, pouring a refill and the timing of a repeat purchase can each change whether a seemingly simple product is useful. I want those circumstances inside the product decision before the range is presented as universal. The work begins with the differences people show us, including differences that make our preferred proposition less tidy.

At the fictional Belvarel Retail in Sydney, I lead 138 product colleagues through six direct reports with A$28.4 million delegated FY 2026–27 expense. The company reported A$1.85 billion revenue across 180 stores in FY 2025–26 and has 4200 employees at October 2026. Buying commitments, store operations, quality and safety assurance, supply chain, finance and technology remain separately led. Those company figures provide context; they do not describe my personal authority over the whole retailer.

My career began in homewares analysis in 1994. It includes instructions I approved that participants misread, a refill range that retained two useful formats and a furniture-cover programme that gave up an unsupported universal-fit claim. The examples here are invented. Their purpose is to show the product choices, the colleagues who challenged them and the limits of the evidence behind a decision.

Three lid footprints still needed eighteen container formats

The 2023 storage study visited sixty households: thirty-seven had shallow cupboards and twenty-three had deeper ones. A$4.6 million supported a proposition of eighteen formats around three lid footprints, replacing a proposed twenty-eight-format range. In a subsequent 120-household pilot, eighty-four accepted the proposition and thirty-six retained unmet needs. That bounded result did not establish a universal storage solution or a national market share.

The drawing I called clear was not clear to eighteen participants

In 2006, eighteen of forty-eight participants reversed a stool cross brace using perspective-only instructions I had approved. Owen Calverden challenged my explanation. A$280000 funded a revised sequence and labels. A different forty-person exercise produced thirty-four completed assemblies and six needing help, keeping the unresolved cases visible rather than declaring the instructions successful for everyone.

A recurring delivery did not imply a common household rhythm

After twelve weeks in the 2025 replenishment trial, sixty-two of 180 households wanted the same four-week interval, seventy-eight wanted variable timing and forty wanted manual ordering. A$5.1 million supported revising the proposition around choice and pause. Trial participation was not treated as a count of paying subscribers, and no retention or revenue result was claimed.

Current focus

I am examining how a product team recognises the limits of a convenient default. A fixed interval, one cover pattern or a standard container depth can make a range easier to explain while leaving important use situations outside it. I want research to reveal those situations early enough to alter the proposition. The question is not whether every individual preference can become a separate product; it is which observed differences change the usefulness of the choice we ask people to make.

Conversations I welcome

I would welcome a conversation about household product decisions, research that changes a range or the contribution a product leader could make to a first board appointment. Any external role would depend on employer agreement, preparation and a realistic time assessment. This invented profile offers no retail products or product-safety advice. An initial message should describe the decision and intended discussion without sharing identifiable research participants, customer records or unreleased range material; appropriate permissions can be established before further evidence is considered.

01

Household context before a standardised range

Explore the decision record

02

Instructions examined through the task people actually perform

Explore the decision record

03

Useful format choices with their limits explained

Explore the decision record

04

Repeat-use propositions that respect different routines

Explore the decision record

138

Product colleagues

Six direct reports at October 2026; buying and store operations separately led

A$28.4 M

Delegated product expense

FY 2026–27 function scope, distinct from company revenue and purchasing commitments

32

Years around household use

Career began February 1994; executive board presentations since July 2021

I believe

I want product development to begin with the cupboard, the hand, the room and the routine. The person using a household object should not have to reproduce the conditions of our presentation to make it useful. A smaller range can make choice easier, but only if the remaining choices still serve the differences we have observed. I would rather explain those differences than hide them inside an elegant universal proposition.

— Phoebe Ellerwick

The work

Decisions that moved the needle.

  1. 01

    The storage range became smaller without assuming every cupboard was the same

    The 2023 storage proposal at the fictional Belvarel Retail began with twenty-eight container formats and an ambition to present one easily understood modular system. I initially favoured a strongly standardised set. Mira Vensley challenged the assumed cupboard depth using visits to sixty households: thirty-seven had shallow cupboards and twenty-three had deeper ones. The important finding was not that those proportions described Australia. It was that the proposed system encountered materially different spaces within the households we had actually observed. A container could belong neatly to our range while failing to fit the place where someone needed it. The commercial team wanted fewer formats to explain and replenish. That was a legitimate concern; retaining every existing variation would have avoided the design decision rather than solved it. Mira wanted the shallow-cupboard situations represented before we reduced the range. I asked the team to distinguish common elements from dimensions that needed to remain different. Sharing a lid footprint could simplify use without requiring every container to have the same depth. We selected eighteen formats organised around three common lid footprints and supported A$4.6 million of development work. The choice gave up the cleanest version of one universal set. It also meant that buying and supply colleagues still had several formats to manage. In a subsequent 120-household pilot, eighty-four accepted the proposition and thirty-six retained unmet needs. We kept those thirty-six in the review rather than describing them as irrelevant exceptions to an otherwise finished solution. The pilot supplied an account of the response to this proposition in this group. It did not establish national demand, a sales forecast or a successful public launch. My contribution was to connect the simplification decision to the observed space around the product. I did not own all buying commitments or certify every physical use. The consequence was a more qualified range proposal with a visible unresolved group, rather than a universal claim built on a convenient cupboard assumption. I learned that modularity should be assessed where the modules will be used. A common component earns its place by helping the person choosing and using the object, not merely by making the presentation of the range more symmetrical.

  2. 02

    I had approved an assembly drawing whose perspective hid the brace orientation

    In 2006, while I was product manager at the fictional Cervellan Homewares, I approved a perspective-only instruction drawing for a flat-pack stool. During an observed exercise, thirty of forty-eight participants completed the sample assembly while eighteen reversed a cross brace. I had treated the drawing as self-explanatory because I already knew how the parts belonged together. Owen Calverden watched the task and challenged my suggestion that printing the same image larger would be enough. The problem concerned interpretation of orientation and sequence, not simply the size of the illustration. I wanted to keep the existing artwork and the development timetable. Owen argued that the instruction needed to show the decision a first-time assembler actually faced. We returned to the point where the brace could be read in two ways and changed the orientation shown in the sequence. We added part labels and examined how the reader would move from one step to the next. A$280000 supported the illustration and instructional-development work. I accepted the extra work rather than asking a general caution at the end to repair a confusing choice earlier in the process. The retest used forty different participants. Thirty-four completed the sample assembly and six needed assistance. Those results were useful, but they did not prove that the new sequence worked unaided for everyone. The different groups also made it inappropriate to present a precise population improvement from a simple subtraction. Quality and engineering colleagues retained responsibility for the assessments within their remit. Neither exercise was described as a safety certification, and the case asserts no injury, recall or regulatory finding. The mistake was mine: I had approved an account of the task that relied on knowledge the reader did not yet possess. The correction required me to withdraw that assumption in front of colleagues who had acted on it. I could preserve the useful product-development work without defending the original drawing. The consequence was a revised sequence with an explicitly unresolved assistance group. Since then, I have wanted instruction reviews to show what a person does at the confusing step, rather than rely on the confidence of people who already know the answer.

  3. 03

    The refill proposition retained a smaller pourable choice alongside bulk

    A 2012 refill-format review at the fictional Belvarel Retail asked seventy-two households how they would use the proposed household range. Forty-four preferred a smaller pourable format and twenty-eight preferred bulk. The buying sponsor favoured a bulk-only proposition because a single format appeared simpler to stock and communicate. I understood the operating attraction, but the research described two different use situations. Asking the first group to accept bulk would not automatically make their handling and storage circumstances disappear. The product decision needed to explain what the simpler offer would give up. I asked the team to examine the reason for each preference rather than turn the counts into a contest with one winner. The smaller format suited people who wanted an easier pouring and storage task; the bulk format served a different purchase and use pattern. The study was limited to the households involved and did not justify a national share estimate. It nevertheless gave us enough specific situations to reject the assumption that bulk was an equally useful choice for everyone. We supported A$1.6 million development of two formats. That decision retained additional range and replenishment work. The sponsor's concern remained valid, so buying and supply colleagues needed to assess the consequences alongside product development. In a later ninety-household pilot, fifty-two selected the smaller format and thirty-eight selected bulk. We described those choices as pilot responses to the proposed alternatives. We did not call them a forecast, a realised retail result or proof that one format should eventually replace the other. No environmental saving followed from these observations alone, and none is claimed here. The result was a proposition with two deliberately different choices and an explanation of why both remained. I was responsible for the product recommendation and its development expense, not all purchasing or retail operating decisions. The trade-off was more complexity for the range in return for serving use situations the bulk-only proposal would have excluded. This case taught me to ask what simplicity means for the household as well as the retailer. One line on our planning sheet can still create an awkward task in someone's kitchen or utility cupboard.

  4. 04

    Universal fit became a twelve-pattern development question

    In 2018, the fictional Belvarel Retail reviewed removable covers for thirty-six furniture frames. The existing cover pattern fitted twenty-four; twelve required different patterns. A commercial colleague wanted to keep a universal-fit description because it made the range easy to explain. I opposed extending that phrase across frames the pattern did not serve. The gap concerned the physical relationship between the cover and the frame, not merely a preference about colour or presentation. A broad label would have transferred an unresolved fit problem from our development meeting to the customer's room. The alternative was not to pretend every frame required a unique new programme. The twenty-four already served by the pattern were a useful common group. I wanted the twelve remaining frames identified accurately and the development choice made in relation to them. We supported A$3.2 million for a twelve-pattern programme, with measurement, prototyping and quality work retaining their respective owners. We also required the eventual fit guide to describe the relationship between the available pattern and the frame, rather than make the customer infer it from a universal promise. The programme created more pattern work and a less sweeping commercial message. I accepted those consequences. Funding the programme did not establish that every new pattern had already been released or that every household could fit a cover without difficulty. The account also did not show that furniture lasted longer, that customers saved money over a lifetime or that the range produced an environmental benefit. Those would require other evidence. The immediate result was a defined development scope and withdrawal of a description broader than the fit assessment supported. My product remit allowed me to shape that scope and the proposition, while buying, quality and customer-service responsibilities remained separately held. The case made a useful distinction between a common range identity and universal physical compatibility. A family of products can still have different fit conditions. The leadership decision is to keep those conditions understandable when the range reaches the person choosing it. I would rather have a customer recognise the correct relationship in a slightly fuller guide than discover the exception after accepting an elegant but unsupported promise.

  5. 05

    The replenishment trial did not turn every participating home into a four-week subscriber

    The 2025 replenishment trial at the fictional Belvarel Retail invited 180 households to examine a proposed automatic four-week delivery rhythm. After twelve weeks, sixty-two wanted the same interval, seventy-eight wanted variable timing and forty preferred manual ordering. The service sponsor wanted to retain one rhythm because it simplified the operating proposition. I saw the attraction, but the responses described different household routines. Participation in the trial did not mean that every household wanted an ongoing subscription, still less that every participant was a paying customer whose retention had been measured. I asked the product discussion to keep the three preferences visible. The fixed interval served one group, but it should not become the condition under which the others could access the proposition. We supported A$5.1 million of work to revise the design around an explicit choice of timing and the ability to pause. The forty households preferring manual ordering remained part of the account. The decision gave up the cleanest version of a universal recurring-service story and introduced more choices for the teams designing and operating the proposition. This was a product-development decision, not a statement that a charging flow had been implemented or that every commercial and legal condition had been resolved. Technology, retail operations and the relevant specialist owners needed their own assessments before a real service could proceed. The fictional evidence contains no subsequent subscriber count, revenue result or retention rate. It also does not prove that the preferences would remain unchanged beyond the twelve-week observation. The immediate consequence was a revised proposition that no longer required every household to share the sponsor's preferred calendar. I learned to treat a regular purchase as a question about use rather than a ready-made service category. A product can be consumed repeatedly while the useful ordering moment varies. A default helps when it reduces effort for someone whose situation fits it; it becomes a poor promise when other routines must be explained away. My responsibility was to make those differences consequential in the design decision. The next stage would need to establish whether the revised choices were understandable and workable, rather than announce that research participation itself had already created a durable commercial relationship.

Others say

Word from the boardroom.

Phoebe accepted that my cupboard observations changed the storage proposition. She kept the unmet situations visible even when a simpler range story would have been easier to present.
Mira VensleyFictional household-research colleague; counterpart in the 2023 storage study
I showed Phoebe how participants interpreted the brace drawing she had approved. She changed the sequence and kept responsibility for the original assumption with herself.
Owen CalverdenFictional product-use colleague; counterpart in the 2006 assembly-instruction review

Point of view

On the record.

An invitation

Contact

I would welcome a conversation about household product decisions, research that changes a range or the contribution a product leader could make to a first board appointment. Any external role would depend on employer agreement, preparation and a realistic time assessment. This invented profile offers no retail products or product-safety advice. An initial message should describe the decision and intended discussion without sharing identifiable research participants, customer records or unreleased range material; appropriate permissions can be established before further evidence is considered.

Open to
  • Board appointments
  • Speaking invitations
Write to Phoebe

office@phoebeellerwick.example